Genetic polymorphisms are well recognized mechanisms underlying inter-individual differences in disease risk as well as treatment response in humans (Evans and Relling (1999) Science 286:487-491; Shields and Harris (2000) J. Clin. Onc. 18:2309-2316). Single nucleotide polymorphism (SNP) analysis has drawn much attention with the hope of identifying genetic markers for and genes involved in common diseases because of the frequency of the SNPs. Also, for many genes, the detection of SNPs known to confer loss of function provides a simple molecular diagnostic to select optimal medications and dosages for individual patients (Evans and Relling (1999) Science 286:487-491). It is common for genes to contain multiple SNPs, with haplotype structure being the principal determinant of phenotypic consequences (Collins et al. (1997) Science 278, 1580-81; Drysdale et al. (2000) Proc. Natl. Acad. Sci. 97:10483-8; Krynetski and Evans (1998) Am. J. Hum. Gen. 63:11-16). Therefore, to more accurately associate disease risks and pharmacogenomic traits with genetic polymorphisms, reliable methods are needed to unambiguously determine haplotype structure for multiple SNPs or other nucleic acid polymorphisms or mutations within genes as well as non-coding genomic regions.
However, current genotyping technologies are only able to determine each polymorphism, including SNPs, separately. In other words, there is a lack of information on how several polymorphisms are associated with each other physically on a chromosome. A DNA haplotype, the phase determined association of several polymorphic markers (e.g., SNPs) is a statistically much more powerful method for disease association studies. Yet unfortunately, it is also much harder to determine a haplotype. Current experimental approaches include a physical separation of homologous chromosomes via means of mouse cell line hybrid, cloning into a plasmid and allele specific PCR. Neither of them is simple enough a method for routine high-throughput analysis. There are also ways to computationally determine haplotypes, but the accuracy of such computational analysis is uncertain.
Approaches that can be used to haplotype SNPs or other nucleic acid polymorphisms, modifications and/or mutations that reside within relatively close proximity include, but are not limited to, single-strand conformational polymorphism (SSCP) analysis (Orita et al. (1989) Proc. Natl. Acad. Sci. USA 86:2766-2770), heteroduplex analysis (Prior et al. (1995) Hum. Mutat. 5:263-268), oligonucleotide ligation (Nickerson et al. (1990) Proc. Natl. Acad. Sci. USA 87:8923-8927) and hybridization assays (Conner et al. (1983) Proc. Natl. Acad. Sci. USA 80:278-282). A major drawback to these procedures is that they are limited to polymorphism detection along short segments of DNA and typically require stringent reaction conditions and/or labeling. Traditional Taq polymerase PCR-based strategies, such as PCR-RFLP, allele-specific amplification (ASA) (Ruano and Kidd (1989) Nucleic Acids Res. 17:8392), single-molecule dilution (SMD) (Ruano et al. (1990) Proc. Natl. Acad. Sci. USA 87:6296-6300), and coupled amplification and sequencing (CAS) (Ruano and Kidd (1991) Nucleic Acids Res. 19:6877-6882), are easily performed and highly sensitive, but these methods are also limited to haplotyping SNPs along short DNA segments (<1 kb) (Michalatos-Beloin et al. (1996) Nucleic Acids Res. 24:4841-4843; Barnes (1994) Proc. Natl. Acad. Sci. USA 91:5695-5699; Ruano and Kidd (1991) Nucleic Acids Res. 19:6877-6882).
Long-range PCR (LR-PCR) offers the potential to haplotype SNPs that are separated by kilobase lengths of genomic DNA. LR-PCR products are commonly genotyped for SNPs, and haplotypes inferred using mathematical approaches (e.g., Clark's algorithm (Clark (1990) Mol. Biol. Evol. 7:111-122). However, inferring haplotypes in this manner does not yield unambiguous haplotype assignment when individuals are heterozygous at two or more loci (Hodge et al. (1999) Nature Genet. 21:360-361). Physically separating alleles by cloning, followed by sequencing, eliminates any ambiguity, but this method is laborious and expensive. Long-range allele-specific amplification negates both of these problems, but is limited to SNP-containing alleles that have heterozygous insertion/deletion anchors for PCR primers (Michalatos-Beloin et al. (1996) Nucleic Acids Res. 24:4841-4843). More complex technologies have also been used, such as monoallelic mutation analysis (MAMA) (Papadopoulos et al. (1995) Nature Genet. 11:99-102) and carbon nanotube probes (Woolley et al. (2000) Nature Biotech. 18:760-763), but these are either time consuming (MAMA), or require technology that is not widely available (nanotubes). U.S. Patent Application No. US 2002/0081598 discloses a haplotyping method which involves the use of PCR amplification and DNA ligation to bring the polymorphic nucleic acid sites in a particular allele into close proximity to facilitate the determination of haplotypes spanning kilobase distances. However, this method relies on at least two enzymatic steps to create DNA fragments that can be ligated with other DNA fragments, and subsequently ligases to combine the DNA fragments to form one large fragment with several polymorphic sites in a shorter distance. These additional sample preparation steps make large scale use and automation of this technique cumbersome and error prone.
Haplotypes, combinations of several phase-determined polymorphic markers in a chromosome, are extremely valuable for studies like disease association1,2 and chromosome evolution. Direct molecular haplotyping has relied heavily on family data, but is limited to short genomic regions (a few kilobases). Statistical estimation of haplotype frequencies can be inconclusive and inaccurate3.
With the rapid discovery and validation of several million single nucleotide polymorphisms (SNP), it is now increasingly practical to use genome-wide scanning to find genes associated with common diseases1,2. However, individual SNPs have limited statistical power for locating disease susceptibility genes. Haplotypes can provide additional statistical power in the mapping of disease genes4-7.
Haplotype determination of several markers for a diploid cell is complicated since conventional genotyping techniques cannot determine the phases of several different markers. For example, a genomic region with three heterozygous markers can yield 8 possible haplotypes. This ambiguity can, in some cases, be solved if pedigree genotypes are available. However, even for a haplotype of only 3 markers, genotypes of father-mother-offspring trios can fail to yield offspring haplotypes up to 24% of the time. Computational algorithms such as expectation-maximization (EM), subtraction and PHASE are used for statistical estimation of haplotypes4,8,9. However, these computational methods have serious limitations in accuracy, number of markers and genomic DNA length. For example, for a haplotype of only 3 markers from doubly heterozygous individuals, the error rates of the EM and PHASE methods for haplotype reconstruction can be as high as 27% and 19%, respectively3. Alternatively, direct molecular haplotyping can be used based on the physical separation of two homologous genomic DNAs prior to genotyping. DNA cloning, somatic cell hybrid construction, allele specific PCR and single molecule PCR10-12 have been used, and these approaches are largely independent of pedigree information. These methods are limited to short genomic regions (allele-specific PCR and single molecule PCR) and are prone to errors.
Therefore, a simple and more reliable method, which is also suitable for large scale and automated haplotype determination of several polymorphic alleles separated by several kilobase distances is needed to facilitate the analysis of haplotype structure in pharmacogenomic, disease pathogenesis, and molecular epidemiological studies.